How srisailesh3320/resai works
Project type and purpose
• This is a single‐page, full‐stack (monolithic) Python application with a Streamlit “frontend” that drives a simple “backend” pipeline.
• Purpose: Automate resume screening by extracting text, invoking an LLM to identify skills, matching those skills to a job description, ranking candidates, and displaying results in a web UI.
File structure
• Top level:
• main.py → Streamlit app (UI + orchestration)
• text_extractor.py → PDF & image OCR/text‐parsing logic
• skill_extractor.py → LLM API calls for skill identification
• resume_analyzer.py → Skill‐matching and ranking logic
• config.py → Paths, environment variables, API key management
• requirements.txt / .env / .gitignore / .devcontainer → setup & config
• No separate “frontend” or “backend” folders, but logical separation by module.
README for additional insights
• Frontend: Streamlit (main.py)
• LLM backend: OpenRouter.ai, Qwen‐3.1b
• Text extraction: PyPDF, pytesseract, EasyOCR, Pillow
• Data processing: pandas, scikit‐learn
• The pipeline is: Upload → Extract text → Call LLM → Compare skills → Display results
Final recommendations
• Every module, even if small, to communicate the precise pipeline.
• Don’t omit config.py and the .env → these are critical for API key management.
• Review with stakeholders to confirm nothing is missing before finalizing.